Subject: Re: Why software patents are bad
From: "Karsten M. Self" <>
Date: Fri, 24 Sep 1999 17:56:03 +0000

"Stephen J. Turnbull" wrote:
> >>>>> "kms" == Karsten M Self <> writes:
>     kms> "Stephen J. Turnbull" wrote:
>     >> Except that software covered by a "poison-pill" is not free any
>     >> more.  Double-plus-ungood.
>     >>
>     >> For it to work, you _must_ shoot innocent bystanders:
>     >> programmers and users whose only sin is prior employment at a
>     >> company whose legal department decides to take action against a
>     >> covered project.  This is
>     kms> I don't see that this is the case.
> I know; if you did, you would never have proposed it.

It wasn't clear to me what your "guilt by prior association" term
meant.  If, as I understand it now, you mean that all employees (and
other directly employed|contracted agents) of the company, who aquire
license to the free software through the employer and for the purpose of
doing work for the employer, are affected by the employer's patent
infringement actions (the Legal/Payroll example you give below), this is
_exactly_ the sort of thing I'm talking about.  Yes, they've got to be
shot, but I don't consider them to be innocent bystanders.

You appear to see this as a personal problem affecting employees as
individuals.  I see it as a corporate problem affecting a company
(organization, institution, etc.) as a whole.

Frankly I don't understand your POV on this.  Yes, it's a pain for the
company (and people) involved, but that's sort of the point -- to raise
the consequences of patent infringement activities against free

> The point is that Conglobulation Inc's decision to act on an alleged
> patent infringement is going to be made by the legal department and
> the top executive suite: how many Emacs users are there?  None, I bet.

 < * snip * >

> But over in the Personnel Department there are two guys who just
> managed to get the department head to sign off on a Linux/Apache based

 < * snip * >

> To inflict any pain on Conglobulation Inc, you have to destroy the
> work done in the Personnel Department.  But those people are
> guilty[sic] of patent enforcement only by prior association with the
> corporation.  I really really don't think that's the kind of violation
> of license envisioned in RMS's definition of free software.

My question is:  who is the licensee?  Conglobulation Inc. or the
individual employees.  For proprietary EULA software I use software
licensed to my employer (or client, as a contractor).  Why should this
be different for free software?  For purposes of the license, then, why
shouldn't actions of the licensee (Conglobulation Inc) reflect on its
compliance|noncompliance status with respect to licensing terms?
> In other words, the poison-pill clause is simply hostage-taking.
> _You_ may think of the hostages as "merely" doing the corporation's
> work for pay, 


> and they can of course still use Linux at home.  


> _They_
> will not like being sacrifices for the cause, having their work
> destroyed by "friendly fire"!  

Truth or consequences, Steve.  What's wrong with that?  What we *have*
done is to create a constituency within Conglobulation Inc. which can
argue that CI's patent enforcement policies are causing real harm to the

I also disagree with your evaluation of the scope and breadth of free
software.  Because of the cost, adaptability, and utility of the stuff,
I expect it to become basic infrastructure in much of computing,
networking, and communications -- an essential facility, if you will. 
CI will find that when it comes down to it, they simply won't have a
choice between patents and free software, it will be patents or
continuing business operations in an internet age.  This will be the
case for virtually *any* CI of more than a few score employees operating
in the US, EU, or Asia -- certainly the bulk of the software patent
holders we're concerned with.

> How about independent consultants trying to provide free software
> solutions to large corporations, any of which may very well have some
> software patents they might like to enforce?  If they have good
> lawyers, they'll get paid for the project even though they don't
> actually deliver it---because it is now illegal for the large
> corporation to accept delivery.  OTOH, the large corporation may
> decide to try to break the contract on the grounds that the consultant
> can't deliver an unencumbered product.  Oops.  Better have a _very_
> good lawyer.

Fair point.  Though this would probably be pretty well hashed out after
the first couple of times it occured.  I'd expect to see a contracting
agreement which doesn't hold the contractor liable for encumberances on
use of a product which result from the client's actions WRT the license
for the product.  Hell, you could say the same thing for EULAs of
component software currently which may have all sorts of obscure little
conditions.  If anything, I expect free software licensing to be much
more uniform -- one license or a small handful, rather than hundreds or
thousands, with terms which are generally well understood, at least
within the business and legal communities concened with the area.  This
is a straw man, IMO.
>     kms>  - Hold software licensees responsible for the actions of
>     kms> major patent licensors.  Something like "the friend of my
>     kms> enemy is my enemy" or "gains by poisoned fruit shall poison".
>     kms> I'm not sure this is necessary, possible, or attractive.  I'm
>     kms> curious as to why Stephen suggests it.
> I don't know whether it's possible, but it seems plausible.  

I thought this was what you'd meant initially.  I see it as a
possibility now, but hadn't envisioned it originally.  I don't see it as
necessary, or particularly desireable.

The *only* situation in which I feel it *might* be necessary is the
instance in which a company manages to divorce its product and
technology (patent) holdings.  The previously discussed "corporate veil
piercing" methods should probably cover this case, however.  If the
association is sufficiently weak between product and tech, the company
risks a buyout of their technology -- presumably not an acceptable risk.

Karsten M. Self (
    What part of "Gestalt" don't you understand?

SAS for Linux:
Mailing list:  "subscribe sas-linux" to    
 10:31am  up 1 day, 13:32,  0 users,  load average: 0.54, 0.30, 0.15